DESCRIPTION (adapted from the Abstract): Social and sexual relationships are often structured by characteristics such as age, race/ethnicity, educational background, and religious affiliation. These data are of consequence due to the likelihood of marriage and/or parenthood among sexual partners as well as due to the influence that sexual partnering can have on the spread of infections through a population. The specific aim of this project is to examine the characteristics of heterosexual partners of American adolescents. The project will also examine the consequences of differences in characteristics of heterosexual partners of American adolescents and adults for contraceptive use, experience with sexually- transmitted diseases, sexual practices, concurrency of relationships, and use of alcohol. Finally, the consistency of reporting of partner characteristics and sexual behavior within a partnership will be evaluated for adolescents. The project will use two large national samples: (1) the ADD Health Study, a national sample of adolescents conducted by the University of North Carolina, and (2) the National Health and Social Life Survey (NHSLS), a survey of adults conducted by the University of Chicago. Statistical methods include coefficient kappa and logistic regressions. The SUDAAN statistical package will be used to the clustered nature of the sample design.